Nuclear magnetic resonance investigation of cobalt (II

Jan 1, 1977 - O. Danny Fox, Jenny F.-Y. Leung, Jesse M. Hunter, N. Kent Dalley, and Roger G. Harrison. Inorganic Chemistry 2000 39 (4), 783-790...
0 downloads 0 Views 785KB Size
120 Inorganic Chemistry, Vol. 16, No. 1, 1977

A. G . Sykes were greatly appreciated. Registry No. C O ( N H ~ ) ~ O H 14403-82-8; ~~+, Co(NH3)50HZC, 16632-75-0; Co(NH3)5Mo04+, 60464-07-5; HMoO4-, 14259-84-8.

References and Notes (1) M. Woods and J. C. Sullivan, Inorg. Chem, 12, 1459 (1973). (2) G. P. Haight, Inorg. Chem., 12, 1461 (1973). (3) R. K. Wharton, R. S. Taylor, and A. G. Sykes, Inorg. Chem., 14, 33 (1975). (4) K. Kustin and S. T. Liu, J . Am. Chem. SOC.,95, 2487 (1973). (5) P. F. Knowles and H. Diebler, Trans. Faraday SOC.,64, 977 (1968). (6) H. Diebler and R. E. Timms, J . Chem. SOC.A , 273 (1971). (7) R. Coomber and W. P. Griffith, J . Chem. SOC.A , 1128 (1968). (8) D. S. Honig and K. Kustin, Inorg. Chem., 11, 1 (1972). (9) R. S. Taylor, Ph.D. Thesis, University of Leeds, 1970. (10) F. Basolo and R. K. Murmann, Inorg. Synth., 4, 171 (1953).

D. S.Everhart and R. F. Evilia M. B. Stevenson, R. D. Mast, and A. G. Sykes, J . Chem. SOC.A , 937 (1969). R. H. Moore and R. K. Ziegler, Report LA 2367 and Addenda, Los Alamos Scientific Laboratory, Los Alamos, N.Mex., 1959. H. A. Benesi and J. H. Hildebrand, J . Am. Chem. SOC.,71,2703 (1949). L. G. Sillen and A. E. Martell, Chern.SOC.,Spec. Publ., No 17,55 (1964). R. C. Splinter, S. J. Harris, and R. S. Tobias, Inorg. Chem., 7, 897 (1968). W. Lyness and P. Hemmes, J . Inorg. Nucl. Chem. 35, 1392 (1973). F. A. Posey and H. Taube, J . Am. Chem. SOC.,78, 15 (1956). D. S. Honig and K. Kustin, J . Am. Chem. SOC.,95, 5525 (1973). J. Aveston, E. W. Anacker, and J. S. Johnson, Inorg. Chem., 3,735 (1964). D. N. Hague and M. Eigen, Trans. Faraday SOC.,62, 1236 (1966). W. A. Johnson and R. G. Wilkins, Inorg. Chem., 9 , 1917 (1970). E. F. C. H. Rohwer and J. Cruywagen, J . S. Afr. Chem. Inst., 16, 26 (1963); 17, 145 (1964); 22, 198 (1969). J. Cruywagen and E. F. C. H. Rohwer, Inorg. Chem., 14, 3136 (1975). See, for example, Y . Sasaki and L. G. Sillen, Ark. Kemi, 29, 253 (1968).

Contribution from the Department of Chemistry, University of New Orleans, New Orleans, Louisiana 70122

Nuclear Magnetic Resonance Investigation of Cobalt(I1) Aminocarboxylates: Evidence for Pentadentate Coordination by EDTA D. S. EVERHART and R. F. EVILIA* Received June 9, I976

AIC60428G

The nuclear magnetic resonance temperature dependence of the complexes CoL1[EDTAl2-,Co"[ 1,2-PDTA]2-, CoII[EDDA], COII[(EDTA)(CN)]~-, Co11[ED3A]-, CoI1[CyDTAI2-, Co1I[1,3-PDTAl2-, Cor1[(EDTA)(OH)I3-, ColIIIDA], and C O I I ( I D A ) ~are ~ - reported. Racemization is rapid for CoI1[EDTAI2-, Colr[(EDTA)(CS)l3-, CoI1[(EDTA)(OH)l3-, and Co1I[l,3-PDTAI2- at 34 OC; racemization of CoI1[ED3A]- is intermediate at 34 OC and rapid by 50 OC. CoI1[EDTAl2-, Co"[ 1,2-PDTAI2-, and Co"[ 1,3-PDTAI2- exist in solution predominantly as pentacoordinate species. The spectrum of CoI1[EDTAl2-is essentially pD independent in the pD range 2.7-1 1.6; pD dependence above pD 12 results from the formation of a mixed hydroxy complex. Deuteration of CoI1[CyDTAl2-is stereospecific; both in-plane and out-of-plane acetate arms experience first-order base-catalyzed deuteration. First-order rate constants for deuteration of ColIICyDTA] are reported. Cis F= trans equilibration is rapid in COII(IDA)~~at r w m temperature. The reaction ColI(IDA) + CoI1(IDA)2==Co1[(IDA)2 CoI1(IDA) is rapid at elevated temperature with retention of IDA configuration. An sN2 type mechanism is proposed for this reaction. The following abbreviations are used: EDTA4-, (-02CCH2)2SCH2CH2N(CH2C02-)2; 1,2-PDTA4-, (-02CCH2)2NCH(CH3)CH2S(CH2C02-)2; EDDA2-, (-02CCH2)NHCH2CH2NH(CH2C02-); ED3A3-, (-02CC1,3-PDTA4-, (-02CCH2)lH2)NHCH2CH2N(CH2C02-)2; CyDTA4-, ~~UW-~,~-(-O~CCH~)~N(C-HX)N(CH~CO~-)~; NCH2CH2CH2N(CH2C02-)2; IDA2-, (-02CCH2)2NH.

+

Introduction The solution structure of bivalent and trivalent transition metal complexes of EDTA and related ligands has long been investigated. The results of these investigations are conflicting and no one has succeeded to date in unambiguously determining the number of coordination sites actually utilized by EDTA type ligands to coordinate labile metal ions. The early works of Schwarzenbachl concluded that transition metalEDTA complexes assume a pentacoordinate structure with a molecule of H 2 0 , OH-, or some other unidentate ligand occupying the sixth position. Jorgenseq2 on the basis of electronic spectroscopy data, also assigned pentacoordination to such complexes. Higginson3 compared acid dissociation constants of various metal-EDTA complexes and concluded that for most bivalent transition metal-EDTA complexes only five of the possible six sites available are utilized by the EDTA chelate. More recently, Higginson spectrophotometrically determined stability constants for mixed-complex formation of EDTA and various unidentate ligands and concluded that aqueous solutions of metal complexes containing EDTA and other related ligands, PDTA and CyDTA, exist as hexacoordinate species approximately 70% of the time [76% for Co(II)] .4 Charles5 concluded from the viscosities of aqueous solutions containing bivalent metal-EDTA complexes that, with the exception of Pb, a hexacoordinated structure is favored. Later, on the basis of large differences found between the Jones-Dale viscosity coefficients for CoIIEDTA and CoIIIEDTA, Yasuda6 proposed that CoIIEDTA is present as

a pentadentate complex or possibly as a mixture of the pentaand hexacoordinated species. The x-ray studies of have clearly shown that an uncoordinated acetate exists in solid CoIIEDTA but did not allow for the assignment of a coordination number to the solution structure of this complex. The spectrophotometric studies of EDTA complexes mixed with OH- and NH3 led Bhat9 to assume hexacoordination for Co"EDTA and Cu"EDTA and pentacoordination for NiIIEDTA. Burnettlo interpreted the free energy and enthalpy changes which accompanied the protonation of divalent metal-EDTA complexes as a hexacoordinate geometry existing in the original species. Margerum and Rosen" utilized kinetic behavior of NiIIEDTA to assign pentacoordination while Wilkins and Yelin12 interpreted their kinetic scheme to indicate at least 80% hexacoordination for CoIIEDTA. Later, Wilkins and Yelin13 rapidly oxidized the labile CoIIEDTA and examined the resultant inert ColIIEDTA complexes; the results of these experiments were in agreement with their previous conclusions. 13C and I7ONMR spectroscopy led Matwiyoff'j to conclude hexacoordination for NiIIEDTA in a pH range of 4-10 and, at pH ~g the rate of any intramolecular process that interchanges proton conformation is slow on the NMR time scale, four signals are observed in the contact shift s ectrum of Ni11EDTA.20128 Rapid A F?1 A conversion of NiPIEDTA at 78 OC, however, results in the averaging of axial and equatorial ethylenediamine protons and also the averaging of axial in-plane with equatorial out-of-plane acetate protons and axial out-of-plane with equatorial in-plane acetate protons yielding a single unresolved peak for all of these resonances.20 Thus, at high temperature, the four-peak spectrum coaleses to a single peak (actually the single peak consists of three unresolved signals consisting of one peak due to all the ethylenediamine protons, one peak due to the averaged in-plane axial and out-of-plane equatorial acetates, and one peak due to the averaged in-plane equatorial and out-of-plane axial acetates). Clearly, observation of only one resonance for ethylenediamine protons at room temperature in CoIIEDTA implies either chemical equivalence (i.e., a planar nonpuckered ring) or rapid interconversion of axial and equatorial protons of a puckered ring. The existence of such a nonpuckered fivemembered ring cannot be discounted at this point; however, considerable evidence has been accumulated which indicates that such a planar ring is energetically u n f a ~ o r e d . ~These ~?~~?~~

122 Inorganic Chemistry, Vol. 16, No. 1, I977

D. S . Everhart and R. F. Evilia

Table I. Spectral Assignments Complex

Shifta

Assignmentb

EDTA

-128 -1 -148 -121 -90 -1 1 -42 +13 +5 2 +89

2 OP A OAc + 2 I P E OAc 2 O P E OAc + 2 IP A OAc + 4 en 2 OP A OAc 2 IP E OAc 1Een 1 OP E OAc 1 OP E OAc 3 CH, 2 IP A OAc 2 A en

1,2-PDTA

(1DA)t.s

-88

-16

Complex EDDA

-148

-61 -60 -18 EDTA(CN)

-64

-55 -41 EDTA(0H)

-65 -51 -48

,

Co (I DA) Co(IDA) Co(IDA), Co(IDA)

ShifP

CyDTA

Assignmentb 2 A OAc 2 E OAc 2 E en 2Aen 4 OAc 4 en 4 OAc 4 OAc 4 en 4 OAc

-143 -100 -20 -8

2 OP A OAc 2 IP E OAc -5 2 IP A OAc 2 OP E OAc ED3A -125 Same as EDTA -6 2 RP -1 Same as EDTA Uncoord ligand +1 1,3-PDTA -110 2 OP A OAc + 2 IP E OAc +I 2 RP 4 en -55 +14 2 RP 2 O P E OAc + 2 IP A OAc -33 2 RP +19 +lo1 P CH, 2 RP +56 a Ppm from water. The following shorthand is used in the assignments: IP, in-plane; OP, out-of-plane; A, axial; E, equatorial; OAc, acetate; en, ethylenediamine; RP, ring proton. -10

previous studies, along with further evidence presented below, indicate that a rapid averaging of axial and equatorial protons of the ethylenediamine protons is responsible for the single resonance observed. The consequences of nitrogen inversion and A + A conversion have been previously considered in Conformational averaging of "backbone" protons eliminates the possibility of only nitrogen inversion acting as an intramolecular averaging process because, in such a process, the ethylenediamine protons retain their conformational identity. A F= A conversion, however, results in racemization of the complex. If only Fermi contact shifts were operative, rapid A + A conversion would lead to observation of either one acetate peak (as seen in Ni"EDTA) or two closely spaced acetate peaks and one ethylenediamine proton peak. The presence of pseudo contact shifts in the cobalt(I1) complexes, however, provides a mechanism to rationalize the appearance of two very different acetate signals under conditions of rapid A + A conversion. For example, if out-of-plane axial acetate protons have a shift of 61, out-of-plane equatorial acetate protons 62, in-plane axial acetates 63, and in-plane equatorial acetates 64, rapid A F= A conversion results in the appearance of peaks at (61 + 64)/2 and (62 + 63)/2.28 If only Fermi contact shifts are present, these two peaks are very nearly identical (e.g., NiIIEDTA), but the presence of large pseudo contact shifts can remove this degeneracy by making 81 # 83 and 82 # 84. To demonstrate that such a room-temperature racemization does in fact occur it is necessary to examine a case where such a process is not possible or a case in which a preferred conformation leads to incomplete averaging. Under these conditions separate resonances for axial and equatorial ethylenediamine protons would be observed as well as resonances for out-of-plane equatorial, out-of-plane axial, in-plane equatorial, and in-plane-axial acetate protons (i.e., a six-line spectrum is expected). Attempts to demonstrate the rapid racemization of CoIIEDTA by slowing the rate to observable time scales at low temperature were, however, unsuccessful. Spectra were obtained in deuteriomethanol to -30 OC, but excessive broadening made it impossible to tell if the racemization was slowed significantly. It is possible, however, to introduce a conformational preference by incorporation of an alkyl group into the

50

Figure 2. Contact shift spectrum of Co1'(l,2-PDTA) at pD 8.0 and 34 "C (shift scale ppm from H,O).

ethylenediamine portion of the ligand.21 The strong pseudoequatorial preference of the methyl group in Co"[ 1,2PDTA] offers an excellent model for viewing such a desired conformation. Because of this conformational preference, averaging is incomplete even though A + A conversion may be fast.28 Such differentiation of geometrically and conformationally unique acetate protons would facilitate the interpretation of the spectrum of CoIIEDTA because the chemical shift values obtained for this complex should be close to the "frozen" values desired for the Co"EDTA species. Examination of Figure 2 shows that the room-temperature contact shift spectrum of Co"[ 1,2-PDTA] consists of 12 peaks. Under the conditions employed in Figure 3 only 10 resolved peaks are observed. The peaks at -121 and +52 ppm can be resolved into two peaks at different pH's19 and at high temperature. The observation of 12 peaks instead of the six predicted above arises because of the asymmetry introduced by the methyl substituent. Identification of the acetate resonances is easily accomplished by deuteration. Upon deuteration, signals at -148, -121, -71, -42, and +52 ppm disappear indicating that these resonances correspond to acetate protons. Respective intensities of signals at +89, +13, and -90 ppm indicate their assignment as axial ethylenes, methyl, and equatorial ethylene, respectively. It is assumed that the methyl occupies an equatorial environment and, thus, only one equatorial ethylenediamine proton is observed. The assignments of the spectra reported here are summarized in Table I. The effect of the methyl substituent on ethylene-

Inorganic Chemistry, Vol. 16, No. 1, 1977 123

Cobalt(I1) Aminocarboxylates

a

100

6

I

Figure 3. Contact shift spectra of (a) CO'I[(EDTA)(CN)]~-and (b) Co"[(EDTA)(0H)I3- at 34 "C and pD -12 (shift scale ppm from H,O).

diamine protons is clearly seen in the contact shift spectrum by the presence of two signals for axial ethylene protons. A similar effect on the protons of the in-plane and out-of-plane acetate arms adjacent to the methyl group can be expected and in fact is observed; in-plane and out-of-plane acetate protons opposite the methyl experience essentially no effect from the methyl substituent because of their distances from the group, while acetates on the same side as the methyl are significantly altered. Such an effect further differentiates between acetate protons that are either adjacent to or opposite the methyl. Considering chemical shift values of 81 for out-of-plane axial, 62 for out-of-plane equatorial, 83 for in-plane axial, and 84 for in-plane equatorial acetates, rapid A A conversion results in resonance at (81 &)/2 and (82 + 83)/2 corresponding to the averaged positions of these protons. The structural similarity between CoIIEDTA and Col'[ 1,2-PDTA] suggests that the bond lengths and dihedral angles in these complexes should be similar. Thus, the PDTA resonance positions should represent, to a close approximation, the "frozen" positions of A conversion as the the EDTA complex. Considering A operative mechanism for the observed molecular racemization of CoIIEDTA at 34 OC, the values obtained by averaging the shifts of the in-plane equatorial with out-of-plane axial protons, (81 84)/2, and the in-plane axial with out-of-plane equatorial protons, (82 + 83)/2, of CoI1[l,2-PDTA] should parallel those seen in CoIIEDTA. When these average values are determined, -(148 + 121)/2 = -134 ppm and (-56 52)/2 = -2 ppm, it is seen that they do in fact agree reasonably well with the observed chemical shift values of acetate resonances taken from the contact shift spectrum of CoIIEDTA, -128 ppm for (61 84)/2 and -7 ppm for (82 + 83)/2; the averaged value of -2 ppm for the ethylenediamine protons of Co"[ 1,2-PDTA] also agrees well with the observed value of -7 ppm for those respective protons in CoIIEDTA. These data further support the agrument that CoIIEDTA undergoes rapid racemization at 34 "C via rapid A + A conversion. In a previous publication,20the failure to observe rapid axial +equatorial interconversion for Ni"EDDA and Ni"MED3A up to 105 OC was interpreted to imply that an uncoordinated acetate is an important factor in the rapid A + A conversion of NiIIEDTA at 34 0C.20 Considering the effect of an uncoordinated acetate on the rate of racemization of NirlEDTA(H20)2-, the rapid racemization of CoIIEDTA at room temperature raises the question of whether predominant pentacoordination also exists in ~ O I I E D T A . Figure 3 shows the contact shift spectrum of CoII(EDTA)(CN). Deuteration of this complex allows assignment of resonances at -64 and -41 ppm as acetate signals and that at -55 ppm as ethylenediamine protons. The appearance of only three resonances for this clearly pentacoordinated species indicates that rapid racemization will occur when a ligand a m

+

+

+

+

6

160

Figure 4. Contact shift temperature dependence of Co"(ED3A) at pD 8.5 (shift scale ppm from H,O): (a) 34 O C ; (b) 105 "C.

is uncoordinated. In spite of the strong coordination of cyanide, racemization is rapid in this complex which has an uncoordinated carboxylate. This similar racemization behavior is consistent with the EDTA complex also having an uncoordinated acetate. It is not surprising that the values of the shifts change drastically because the electron distribution within the mixed cyanide complex is certainly different from that in the aquo complex. The important feature is that, since A raonly three peaks are observed once again, the A cemization process is still rapid even though the complex now contains a coordinated cyanide. Since the EDTA ligand certainly utilizes only five of the possible six donor groups in Co"(EDTA)(CN), the resemblance of these spectra seems to indicate that only five of the possible six sites are utilized by EDTA in CoIIEDTA. If C0"EDTA was predominafltly hexacoordinate, addition of the cyanide anion would be accompanied by displacement of a cobalt-acetate bond and significant changes in the contact shift spectrum would be expected. This displacement would almost certainly result in the production of additional peaks since the symmetry of the complex would be destroyed. The shift differences in the spectra of these two complexes are explainable due to the presence of the negatively charged cyanide. The temperature dependency of the spectrum of Co"(EDTA)(CN) is essentially identical with that of CoI1(EDTA)(H20). It should be noted that the previous unresolved resonances at -7 ppm in Co"EDTA separate into two distinct resonances after incorporation of the cyanide anion. Furthermore, at high pH, a spectrum very similar to that obtained for CO(EDTA)(CN)~-is obtained (see Figure 3). This is believed to occur by formation of the hydroxy complex above pD 11.6 via the reaction Co(EDTA)(H,O)'-

+ OH-

-+

CO(EDTA)(OH)~-t H,O

The electrostatic effect of the coordinated negative hydroxy group has an effect on the NMR spectrum very similar to the effect of a negative cyanide group but does not appear to change the structure or rate of racemization. Since the colors of the two solutions of Figure 3 are different, it is apparent that they arise from different species. Further strong evidence for pentacoordination in Co'IEDTA is found in the contact shift spectrum of Co"ED3A (Figure 4). Two widely separated resonances are observed with relative intensities of low- and high-field signals of about 1:2. Comparison of the essentially identical contact shift spectra of Co"ED3A and CoIIEDTA suggests a very strong resemblance in their solution structures and dynamics. The room-temperature spectra of both complexes consist of two signals; the higher field signal in CoIIEDTA, however, consists of two unresolved resonances. The shift differences between the downfield and upfield signals is 121 ppm for CoIIEDTA and 118 ppm for Co"ED3A. Deuteration of Co"ED3A

D. S . Everhart and R. F. Evilia

124 Inorganic Chemistry, Vol, 16, No. 1, 1977

133

0

Figure 5 . Contact shift temperature dependence of Co"(EDDA) at pD 6.3 (shift scale ppm from H,O): (a) 34 'C; (b) 105 "C.

resulted in the same behavior as reported for the EDTA complex. With onIy pentacoordination possible for Co"ED3A it can be concluded that the EDTA complex exists predominantly as a pentacoordinate species also. The possibility of rapid pentacoordinate + hexacoordinate equilibration occurring in CoIIEDTA cannot be completely disregarded from this study; however, the strong similarity of the spectral data for obviously pentacoordinate Co(ED3A)- and CO(EDTA)~strongly suggests that any such equilibration would favor the pentacoordinate form by a large amount. Previously it was proposed that the presence of an uncoordinated acetate arm in the corresponding Ni(I1) complexes greatly increases the rate of intramolecular racemization.20 NiIIEDDA and Ni"MED3A were found not to undergo appreciable intramolecular rearrangement up to -105 OC. Examination of the contact shift spectrum of CoIIEDDA (Figure 5) shows that axial and equatorial protons of both ethylenediamine and acetate protons are well resolved up to 105 OC, implying a slow rate of any intramolecular averaging processes. This spectrum is consistent with the assumed stereochemically rigid transcoordination of EDDA. Some averaging process appears to be occurring at higher p D ' ~ ; ~ l however, it is easily argued that CoIIEDDA should racemize at least as rapidly as C0"EDTA if the averaging process does not involve an uncoordinated acetate arm. In fact, it could be argued that C0"EDDA should racemize even more rapidly than CoIIEDTA because of the smaller number of bonds which must be stretched for this process to occur. Addition of an acetate to EDDA (ED3A) results in the coalesence of both acetate and backbone protons. Careful examination of Figure 4, however, shows that the resonance at --2 pprn is relatively broad at room temperature but becomes sharper at high temperatures. This suggests a kinetic process (which is presumed to be the racemization reaction) whose time scale is comparable to the chemical shift difference of the protons. This same reaction is, however, rapid in Co(EDTA) ('i.e., no kinetic broadening is observed) leading to the conclusion that the rate of racemization of this series is Co(EDDA) < Co(ED3A)- < CO(EDTA)~-.This increase in rate of racemization is consistent with the increasing probability of the complex containing an uncoordinated acetate and indicates that Co(ED3A)- may contain some tetracoordinated form. The spectrum of Co"( 1,3-PDTA) is shown in Figure 6. Deuteration results in assignment of resonances at -1 10 and -33 ppm as acetate signals with the remaining signals at -55 and +lo1 ppm corresponding to protons of trimethylenediamine "backbone". The incorporation of a methylene group into the ethylenediamine backbone makes possible the presence of two isomeric forms: skew and hair.^^.^^ Presence and slow interconversions of these two isomers would result in resonances

100

0

Figure 6. Contact shift spectrum of Co1'(l,3-PDTA) at pD 6.0 (shift scale ppm from H20).

Figure 7. Contact shift spectra of Co*I(CyDTA) at 34 "C (a) before and (b) after deuteration (shift scale ppm from H,O).

for axial and equatorial protons on each carbon (Le., four signals would be observed). However, if any process occurred which rapidly equilibrated these isomers, two signals for the backbone protons would then be observed (average of protons on carbons 1 and 3 and average of protons on carbon 2 ) . The methylene signals in the contact shift spectrum of Co"( 1 3 PDTA) indicate that such an equilibration does occur. In view of the previously presented data concerning the intramolecular racemization of the Co(I1) complexes mentioned in this investigation, Figure 6 is interpreted to indicate that A + A conversion of Co111,3-PDTA occurs rapidly at 34 "C. As previously argued, the presence of a pseudo contact shift removes the degeneracy of acetate protons under the conditions of rapid A + A equilibration. This differentiation of in-plane and out-of-plane axial as well as in-plane and out-of-plane equatorial acetates can be observed when examining the contact shift spectrum of Co"CyDTA, Figure 7. The assignments of the four acetate signals are listed in Table I. It has been established for Ni(I1) complexes that in-plane acetate protons undergo deuteration at an extremely slow rate.28 The cyclohexane "backbone" of the CyDTA ligand makes this complex stereochemically rigid, incapable of any intramolecular rearrangement other than that which would occur with rupture of the N-Co bond followed by nitrogen inversion. Therefore, the cyclohexane "backbone" of the C0"CyDTA complex prevents acetate arms from experiencing (on the NMR time scale) any process which would allow an in-plane arm to equilibrate to an out-of-plane position. Thus in-plane acetates are fixed in their position and are incapable of reorientating to an out-of-plane geometry. In Figure 7b, it is observed that all acetate protons of CoIICyDTA

Cobalt(I1) Aminocarboxylates

Inorganic Chemistry, Vol. 16, No. I, 1977 125

Table 11. Rate of Deuteration of CoCyDTA Acetates ~

~~

Peak position

First-order rate const (71 "C), s-'

-143

6.5 8.2 2.5 6.0

-100 -20 -8

X X X X

low4 lo-'

Assignment Out-of-plane axial In-plane equatorial In-plane axial Out-of-plane equatorial

bond-breaking mechanism must be operative in the mono e bis reaction, the conformations of the chelate rings are retained. This indicates an S Ntype ~ mechanism in which the new bond is at least partially formed before the old bond is completely broken. These results, therefore, indicate that the estimate that C O ( E D T A ) ( H ~ O ) oxidizes ~5-12 times slower than Co(EDTA)2- l3 may be incorrect. This estimate was used in order to estimate the amount of pentadentate form which could exist when essentially 100% of hexadentate cobalt(II1) species is obtained upon oxidation. It would appear that the oxidation of the pentadentate isomer is much slower than that of the hexadentate form, hence, yielding the previous results. Acknowledgment. This work was supported by the National Institutes of Health under Grant GM20645. We thank Drs. C. N. Reilley and J. E. Sarneski of the University of North Carolina for some 1,3-PDTA. Registry No. CoI1[EDTAI2-, 14931-83-0; Co"[ 1,2-PDTAI2-, 26425-15-0; Co"[IDA(H20)3], 34387-45-6; Co"[IDA]2*-, 233 1445-6; Co"[ED3A]-, 60645-76-3; CO"[ 1,3-PDTAI2-, 16904-03-3; Co"[EDDA], 29977- 10-4; Co"[EDTA(CN)13-, 60645-77-4; CO"[EDTA(OH)] 3-, 1509 1-47- 1; Co"[CyDTAI2-, 28 161-9 1-3.

100

0

Figure 8. Contact shift temperature dependence of Co"(1DA) with [IDA]/[Co] = 1.5/1 at pD 6.2 (shift scale ppm from H,O): (a) 34 " C ; (b) 80 'C; (c) 105 "C.

exchange with deuterium. This is contrary to what is observed in the corresponding NiIICyDTA. Rate constants for the exchange reaction were determined by a method previously outlined by Erickson et al. and are reported in Table 11. Out-of-plane and axial acetates are assumed to deuterate at the fastest rates and the spectral assignments are made on this basis. These assignments are consistent with the 1,ZPDTA assignments made on the basis of the proposed racemization mechanism. The last piece of evidence which supports the contention that an uncoordinated acetate is an important factor in determining the rate of intramolecular rearrangement is seen in Figure 8. This figure shows the behavior of a solution in which the mole ratio of IDA to Co(I1) is 1.5:l. Peaks assignable to mono and bis complexes are observed, but the cis and trans isomers of the bis complex are not observed. This indicates that cis + trans isomerization is rapid in the bis complex. Rapid cis e trans isomerization has also been observed in Ni(IDA)22-. Despite the rapid cis + trans isomerization, the axial and equatorial protons are not averaged indicating that the individual ring conformations are not averaged (i.e., the complex is not racemized). In this case there is no uncoordinated acetate and, hence, the racemization is slow on the NMR time scale. The most surprising feature of the data presented in Figure 8 is the coalescence of peaks due to mono and bis complexes at high temperature. This equilibration occurs without racemization of the complex. In other words, even though a

References and Notes (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)

(12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24)

(25)

(26) (27)

(28) (29) (30)

(31) (32) (33)

G. Schwarzenbach, Helu. Chim. Acta, 32, 839 (1949). C. K. Jorgensen, Acta Chem. Scand., 9, 1362 (1955). W. C. E. Higginson, J . Chem. Soc., 2761 (1962). W. C. E. Higginson, J . Chenz. SOC.A , 1579 (1970). R. G. Charles, J . Am. Chem. Soc., 78, 3946 (1956). M. Yasuda, Bull. G e m . SOC.Jpn., 41, 139 (1968). J. C. Smith and J. L. Hoard, J . A m . Chem. Soc., 81, 556 (1959). S. Richards, B. Pederson, J. V. Silverton, and J. L. Hoard, Inorg. Chem., 3, 27 (1964). T. R. Bhat and M. Krishnamurthy,J . Inorg. Nucl. Chem., 25, 1147 (1963). A. P. Burnett, G. N. Nancollas, and P. N. Smith, J . A m . Chem. SOC., 91, 4680 (1969). D. W. Margerum andH. M. Rosen, J . Am. Chem. SOC.,89, 1088 (1967). R. G. Wilkins and R. Yelin, J . A m . Chem. Soc., 89, 5496 (1967). R. G. Wilkins and R. E. Yelin, J . A m . Chem. Soc., 92, 1191 (1970). N. A. Matwiyoff and C. E. Strouse, J . Am. Chem. Soc., 92,5222 (1970). R. S. Milner and L. Pratt, Discuss. Faraday Soc., 34, 88 (1962). F. F. L. Ho and C. N. Reilley, Anal. Chem., 41, 1835 (1969). F. F. L. Ho, L. E. Erickson, S. R. Watkins, and C. N. Reilley, Inorg. Chem., 9, 1139 (1970). D. R. Eaton and W. D. Phillips, Adu. Magn. Reson., 1, 103 (1965). L. Pratt and B. B. Smith, Trans. Faraday SOC.,65, 1703 (1969). D. S. Everhart and R. F. Evilia, Inorg. Chem., 14, 2755 (1975). C. R. Hare and C. J. Ballhausen, J . Chem. Phys., 40, 788 (1964). D. C. Young and C. N. Reilley, J . Coord. Chem., 1, 95 (1971). M. W. Grant, H. W. Dodgen, and J. P. Hunt, J . A m . Chem. Soc., 93, 6828 (1971). G. L. Blackmer, R. E. Hamm, and J. I . Legg, J . A m . Chem. SOC.,91, 6632 (1969). N. H. Furman, Ed., "Standard Methods of Chemical Analysis", Vol. 1, 6th ed, Van Nostrand, Princeton, N.J., 1962, p 387. K. Mikkelsen and S. D. Nielsen, J . Phys. Chem., 64, 632 (1960). R. K. Force and J . D. Carr, Anal. Chem., 46, 2049 (1974). L. E. Erickson, D. C. Young, F. F. L. Ho, S. R. Watkins, J. B. Terrill, and C. N. Reilley, Inorg. Chem., IO, 441 (1971). R. D. Gillard and H. M. Irving, Chem. Reu., 65, 603 (1965). A. M. Sargeson, Transition Met. Chem., 3, 303 (1966). D. S. Everhart and R. F. Evilia, unpublished results. J. E. Sarneski and C. N. Reilley, Inorg. Chem., 13, 977 (1974). M. K. Doh, J. Fujita, H. Ogino, and K. Saito, Bull. Chem. SOC.Jpn., 48, 874 (1975).